Corpus ID: 202542486

What do Deep Networks Like to Read?

@article{Pfeiffer2019WhatDD,
  title={What do Deep Networks Like to Read?},
  author={Jonas Pfeiffer and Aishwarya Kamath and Iryna Gurevych and Sebastian Ruder},
  journal={ArXiv},
  year={2019},
  volume={abs/1909.04547}
}
  • Jonas Pfeiffer, Aishwarya Kamath, +1 author Sebastian Ruder
  • Published 2019
  • Computer Science
  • ArXiv
  • Recent research towards understanding neural networks probes models in a top-down manner, but is only able to identify model tendencies that are known a priori. [...] Key Method By fine-tuning an autoencoder with the gradients from a fixed classifier, we are able to extract propensities that characterize different kinds of classifiers in a bottom-up manner.Expand Abstract

    Figures, Tables, and Topics from this paper.

    Citations

    Publications citing this paper.

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 36 REFERENCES

    What do Deep Networks Like to See?

    VIEW 2 EXCERPTS

    Analysis Methods in Neural Language Processing: A Survey

    VIEW 1 EXCERPT

    Annotation Artifacts in Natural Language Inference Data

    VIEW 7 EXCERPTS
    HIGHLY INFLUENTIAL